为准确评估大倾角煤层开采过程中飞矸对人或设备的威胁等级，将工作面飞矸运动过程中具有的总动能作为判定飞矸威胁等级大小的依据，采用Rocfall软件建立大倾角工作面飞矸运动模型。选取飞矸的质量、初速度、工作面长度、煤层倾角、碰撞恢复系数6因素5水平进行标准正交试验，通过正交试验法得到6种因素灵敏度大小的排序。根据排序及工作面实际情况，选取影响飞矸动态威胁等级评估的主要因素为飞矸的质量、速度以及沿工作面运动距离3个动态特征量，采用Genie软件建立动态贝叶斯网络模型，通过仿真得到飞矸沿工作面运动全过程的动态威胁等级概率。仿真结果表明：飞矸威胁等级在飞矸运动初始阶段急剧增加，随着时间片的增加，在中期和末尾阶段威胁等级趋于平稳。为了减轻工作面飞矸动能过大对人或设备造成伤害，可在飞矸运动初始位置以及飞矸威胁等级较高的位置设置防护装置，以达到降低飞矸威胁的目的。

In order to accurately assess the threat level of the flying gangues to people or equipment during the mining of the steeply dipping seam, the total kinetic energy of the flying gangue was used as the basis for determining the magnitude of the threat level, and model of flying gangues movement was establish by Rocfall in the steeply dipping seam. The six factors-five levels including flying gangue’s mass, initial velocity, working face length, dip angle of coal seam and collision recovery coefficient were selected for the standard orthogonal experiment, and the order of the sensitivity of the six factors was obtained by the orthogonal experimental method. According to the sorting and the actual situation of the working face,the major factors affecting the dynamic threat level of the flying gangues are the quality, speed and distance moving along the working face. Dynamic Bayesian network model was set up by adopting the Genie, the dynamic probability of threat level was obtained during the whole process of flying gangues movement by simulation. The simulation results show that the threat level of the flying gangues increases sharply in the initial stage of fly gangue movement. As the time slice increases, the threat level tends to be stable in the middle and end stages. In order to reduce the damage caused by the flying gangues of the working face, the protective devices can be set up at the initial position of flying gangues movement and the position with higher threat level to achieve the purpose of reducing flying gangues threat.

0 引言

1 基于正交试验飞矸威胁因素分析

   1.1 工作面飞矸运动模型的构建

   1.2 影响因素分析

   1.3 确定参数取值 

   1.4 正交试验

2 基于动态贝叶斯网络的威胁等级估计

   2.1 贝叶斯网络建模

   2.2 贝叶斯网络概率

   2.3 动态贝叶斯网络建模

3 结论